Arc flash study and labels field guide for electrical crews

What an arc flash is, and why the study exists

An arc flash is an explosive release of energy when an electrical fault jumps across the air instead of through a conductor. A dropped tool, a loose strand, a failed insulator, and the current finds a path through ionized air at temperatures that run several times hotter than the surface of the sun. The blast carries heat, pressure, molten metal, and a sound wave, and the heat is what burns the worker. The study exists to put a number on that heat before anyone opens the door.

That number is incident energy, the thermal energy that would land on a worker standing at a normal working distance if the arc went off. The arc-flash study calculates it at every applicable piece of gear so the crew knows two things they cannot eyeball: how much energy the gear can throw, and how far back the danger reaches. From those, the program sets the protective equipment and the boundary.

The reason this is a calculated study and not a judgment call is that the hazard is invisible and counterintuitive. The most dangerous gear is often not the highest voltage. A 480 V switchboard with a slow upstream breaker and high available fault current can throw far more energy than a 13.8 kV cubicle that clears in three cycles. You cannot tell by looking, by voltage, or by experience. You calculate it, you label it, and you build the safety program around the labels.

What is incident energy, and what is 1.2 cal/cm2?

Incident energy is the amount of thermal energy a worker would receive on the surface of the skin or the PPE at a set working distance from an arc, expressed in calories per square centimeter (cal/cm2). It is the single output the whole study drives toward, because it is what decides whether a task can hurt you and how much protection it takes.

The number to carry in your head is 1.2 cal/cm2. That is the generally accepted level where the onset of a second-degree burn occurs on bare skin from an arc exposure, and it is the value NFPA 70E uses to draw the arc-flash boundary. Below it, an unprotected worker is unlikely to take a second-degree burn from the arc. Above it, arc-rated protection is required. The number sounds small because it is. A cup of hot coffee will do worse, but it does not arrive at the speed and across the area an arc does.

One trap kills people who half-understand this. An incident energy below 1.2 cal/cm2 does not mean there is no arc-flash hazard. It means arc-rated clothing is not required for the thermal exposure at that distance. The blast pressure, the shrapnel, and the shock hazard are all still there. A low cal/cm2 number is a thermal-PPE result, not a permission slip to treat the gear as safe.

NFPA 70E vs IEEE 1584: which does what?

These two documents get named in the same breath and they do completely different jobs. NFPA 70E is the safety standard, the rulebook for the worker side: the electrical safety program, the risk assessment, the boundaries, the PPE selection, the energized-work permit, and what the label has to say. IEEE 1584 is the calculation method, the engineering recipe for computing how much incident energy a given piece of gear will produce. One tells you the rules; the other tells you the number.

They depend on each other. IEEE 1584 gives you the incident energy and the arc-flash boundary at each bus. NFPA 70E takes those values and turns them into action: what PPE to wear, how close anyone can get, when the work needs a permit, and what goes on the label bolted to the gear. Run only IEEE 1584 and you have numbers nobody is required to act on. Run only NFPA 70E without a study and you are guessing at the hazard the rules are supposed to control.

Keep the bodies straight when you cite them. NFPA 70E is published by the National Fire Protection Association and is the standard for electrical safety in the workplace. IEEE 1584 is the IEEE guide for performing arc-flash hazard calculations. OSHA does not publish either one, but its general-industry electrical rules in 29 CFR 1910 Subpart S are what make compliance enforceable, and OSHA treats NFPA 70E as a recognized way to meet that duty. The adopted edition of each controls; confirm it before you cite a clause.

The inputs an IEEE 1584 calculation needs

IEEE 1584 does not run on voltage alone. It needs a set of inputs that come from the electrical model and the gear itself, and a wrong value in any one of them moves the answer. The method covers three-phase systems from 208 V to 15 kV, and outside that range it does not apply and the engineer falls back to other approaches.

The available bolted fault current comes from the short-circuit study. From it, IEEE 1584 derives the arcing current, which is lower than the bolted value because the arc itself is a resistance. The protective device clearing time at that arcing current comes from the coordination study and the device curves, and it is the input that swings the result hardest. The electrode configuration, whether the conductors are vertical or horizontal and inside a box or in open air, the gap between conductors, the enclosure size, and the working distance round out the model. Get the configuration or the gap wrong and the incident energy is wrong in a way that still looks like a clean calculation.

Why does the clearing time drive the incident energy?

Incident energy is, at its core, power multiplied by time. The arc dumps energy at a rate set by the current and the voltage, and the total a worker absorbs depends on how long the arc burns before something upstream clears it. Double the clearing time and you roughly double the energy. That is why the protective device, not the voltage, is usually the thing that decides whether a job needs a face shield or a full arc-flash suit.

This is the lever that surprises people new to the work. A breaker sitting in its instantaneous region clears in a few cycles and the energy stays low. The same breaker on a long-time delay, or one whose pickup is set above the arcing current so the arc never reaches the fast part of the curve, can let the arc burn for hundreds of milliseconds or longer. The arcing current matters here too: if it lands below the instantaneous threshold, the device times out on the slow curve and the energy climbs fast.

IEEE 1584 caps the arc duration at 2 seconds when no device will clear it sooner, on the reasoning that a worker can usually move away in that time. Treat that cap as a modeling convention, not a comfort. Two seconds of arc on a high-fault-current bus is a catastrophic number, and it is exactly the situation arc-energy reduction is meant to fix. The clearing time is where the study and the coordination work meet, which is why the two studies are done together and verified in commissioning.

The studies that come first: short-circuit and coordination

You cannot do an arc-flash study in isolation. It is the third study in a stack, and the two beneath it are not optional. The short-circuit study establishes the available fault current at every bus. The protective device coordination study establishes how fast each device clears at the fault levels the short-circuit study found. Only with both in hand can IEEE 1584 produce a defensible incident energy.

The order is not a formality. The short-circuit result feeds the coordination work, and both feed the arc flash. Change the utility available fault current, add a transformer, swap a breaker, and the short-circuit numbers move, the coordination shifts, and every incident-energy result downstream is now stale. This is the most common way a study goes wrong: the gear changed and only the labels stayed the same.

Anyone selling an arc-flash study that did not start with a current short-circuit model and a coordination study is selling labels, not a study. The labels will have numbers on them. The numbers will be fiction. On a data center or any plant with high available fault current, this stack is also where the arc-energy reduction strategy gets designed, because the same clearing-time decisions that make the system coordinate are the ones that hold the incident energy down.

What is the arc-flash boundary?

The arc-flash boundary is the distance from the gear at which the incident energy drops to 1.2 cal/cm2, the onset of a second-degree burn. Inside that distance, anyone exposed to the gear needs arc-rated PPE rated for the energy at their working distance. Outside it, the thermal hazard from that arc is below the burn threshold. The study calculates this distance for each piece of gear, and it goes on the label.

Do not confuse the arc-flash boundary with the shock approach boundaries, which are a different hazard. NFPA 70E sets a limited approach boundary and a restricted approach boundary around exposed energized parts, and those are about electric shock, not arc heat. They are not tied to incident energy at all. The limited approach is the distance an unqualified person should not cross without an escort and the proper conditions. The restricted approach is the closer line where shock from arc-over and an inadvertent move becomes likely, and only a qualified person with shock PPE and a plan crosses it.

So a given task can sit inside two different bubbles at once: an arc-flash boundary measured in feet for the heat, and a restricted approach boundary measured in inches for the shock. They are computed separately, they protect against different injuries, and the PPE for one does not cover the other. A worker can be arc-rated head to toe and still be unprotected against the shock if the voltage gloves are wrong.

What goes on an arc-flash label?

The arc-flash label is the study made visible on the gear, and NFPA 70E spells out what it has to carry, commonly cited at 130.5(H) in recent editions. At a minimum the label shows the nominal system voltage, the arc-flash boundary, and at least one piece of PPE-selection information. That last part is where the two methods diverge, and the label has to pick a lane.

The PPE information must be one of: the available incident energy and the working distance it was calculated at, or the arc-flash PPE category for the equipment, but not both on the same label. The label can instead carry the minimum arc rating of clothing or a site-specific PPE level. When the incident-energy method is used, the label also carries the date of the study, because an incident-energy number is only as good as the system model behind it, and the date tells a worker how stale that model might be.

Read the label like the inspector does. Voltage tells you the shock hazard. The arc-flash boundary tells you when to gear up. The incident energy at the working distance, or the category, tells you what to wear. If a label shows a category and an incident energy together, someone mixed the methods, and that label cannot be trusted. The exact clause numbers shift between editions, so confirm the required contents against the adopted edition of NFPA 70E before you order labels.

Incident-energy method vs PPE category method

NFPA 70E gives two ways to pick arc-flash PPE, and they cannot be mixed on the same task or the same piece of gear. The first is the incident-energy analysis method: you use the calculated cal/cm2 at the working distance and select clothing and equipment with an arc rating that exceeds it. The second is the arc-flash PPE category method: you look the equipment up in the standard's tables, which return a category from 1 to 4, and you wear what the category requires. The choice of which method to use, commonly addressed at 130.5(F), is made per task, and the no-mixing rule is firm.

The category method exists for facilities that have not run a full study, and it has hard limits on the parameters it can be applied to, set by the tables. If your gear falls outside those bounds, or the available fault current and clearing time are higher than the table assumes, the category method does not apply and you need the calculation. The incident-energy method is more precise and is what a real study produces, which is why it is the method behind a labeled, studied facility.

Mixing the two is a classic field error that reads as diligence and is actually a hole. A crew sees a category 2 on a table reference and an 8 cal/cm2 number from somewhere else and splits the difference, or reads a label that someone built with both. Pick one method for the task and stay in it. The numbers behind the two methods are not interchangeable, and averaging them gives you PPE that matches nothing.

Arc-rated PPE and the categories

The rule for arc-rated PPE is one sentence: the arc rating of the clothing and equipment, in cal/cm2, has to exceed the incident energy the worker is exposed to at the working distance. Arc rating is the protection number, derived from the arc thermal performance value (ATPV) or the energy of breakopen threshold (EBT), and it is printed on the garment. An incident energy of 8 cal/cm2 needs PPE rated above 8, not at 8.

Note the term. It is arc-rated (AR), not flame-resistant (FR). All arc-rated clothing is flame-resistant, but not all flame-resistant clothing has a tested arc rating, and only the arc rating tells you the cal/cm2 it protects against. A garment that says FR with no arc rating is not arc-flash PPE. This trips up crews who buy on the label color and not the number.

When the category method is used, the categories carry minimum arc ratings: category 1 at 4 cal/cm2, category 2 at 8, category 3 at 25, and category 4 at 40. Each category specifies the full kit, which past category 1 means arc-rated layers, an arc-rated face shield or a hood with a balaclava, and the right gloves, including rubber insulating gloves with leather protectors for the shock hazard. Above 40 cal/cm2, the standard stops giving you a category, because the blast pressure at that energy is its own hazard that clothing does not solve. That is the energy where you stop dressing for the arc and start removing it.

How do you reduce the arc-flash hazard?

PPE is the last line, not the first. NFPA 70E lays out a hierarchy of risk control, commonly cited at 110.5(H), and PPE sits at the bottom of it because it does nothing to lower the hazard, it only tries to survive it. The top of the hierarchy is elimination: put the gear in an electrically safe work condition. De-energized, locked out, tested dead, and grounded where required, the arc-flash hazard is zero. That is the only state with no risk, and it is the answer the standard wants you to reach for first.

When the work genuinely cannot be done dead, you reduce the energy by engineering before you reach for the suit. Faster clearing is the most direct lever, because energy scales with time. An energy-reducing maintenance switch lets a worker put a trip unit into a no-intentional-delay mode while inside the boundary, so a fault clears fast during the exposure, then returns to the coordinated setting afterward. Differential relaying, zone-selective interlocking, current-limiting devices, and arc-energy active mitigation systems all chase the same goal of cutting the time the arc burns.

The NEC backs this for larger gear. NEC 240.87, the arc-energy reduction article, requires a means to reduce clearing time on circuit breakers where the trip is rated or can be set to 1200 A or higher, and lists acceptable methods including the maintenance switch, differential relaying, zone-selective interlocking, an instantaneous trip, and active arc-flash mitigation. The article number and its scope have been tightened across recent code cycles, so confirm the current requirement against the adopted NEC edition. The principle does not change: clear it faster and the energy drops with the time.

The energized electrical work permit

The default is dead. Energized work is the exception, and NFPA 70E only allows it under narrow justifications: when de-energizing would introduce a greater hazard, when it is genuinely infeasible because of the equipment design or the operation, or for diagnostics and testing that cannot be done with the circuit dead. Inconvenience and schedule are not on that list, and that is the line crews push hardest.

When the work clears that bar, it runs under an energized electrical work permit. The permit is not a rubber stamp. It documents the task, the justification, the results of the shock and arc-flash risk assessment, the boundaries, the PPE, the means used to restrict access, and the sign-offs from the people who own the risk. Filling it out honestly is its own filter, because the act of writing down why this cannot be done dead exposes most of the jobs that actually can.

The permit is also the record that protects the crew and the employer after the fact. If a job goes wrong, the permit shows the hazard was assessed and the controls were in place. If there is no permit and no justification for energized work that should have been done dead, the program failed before anyone opened the cover. Verify the panel is dead before you treat it as dead, every time, no matter what the permit says.

How often do you update an arc-flash study?

NFPA 70E says the arc-flash risk assessment, commonly addressed at 130.5, has to be reviewed at intervals not to exceed 5 years, and updated whenever a change to the electrical system could affect the results. The 5-year clock is the outer limit, not a guarantee the study stays valid that long. A system that changes in year two has a stale study in year two, regardless of the calendar.

The word in the standard is reviewed, not recalculated from scratch. The review confirms whether anything changed that moves the inputs: the available fault current, the protective device settings, the equipment, or the utility configuration. If nothing material changed, the review documents that. If something did, the study gets rerun for the affected parts and the labels get reprinted.

The changes that demand an immediate update are the ones crews make all the time and forget to flag. A new or replaced transformer changes the available fault current. A breaker swap or a trip-setting change moves the clearing time. A utility upgrade raises the fault current at the service. New feeders reroute the fault paths. Each of these can move incident energy up, and a label that now understates the hazard is worse than no label, because it tells a worker to gear down for a job that got more dangerous.

The data center lineup: high fault current and the maintenance switch

Data centers concentrate every factor that drives arc-flash energy up. The available fault current is high because the service is large and stiff, transformers are big, and parallel sources and generators feed the bus. High fault current plus a slow device equals a large incident energy, and the gear is exactly where people have to work to maintain a plant that never goes dark.

The design answer is selective coordination plus arc-energy reduction, working together. Selective coordination keeps a fault on one branch from tripping the whole hall, which means upstream devices are intentionally delayed, which pushes the incident energy up at those upstream buses. The maintenance switch on the main and on the larger breakers resolves the conflict: coordinate for selectivity in normal operation, then switch to fast clearing while a worker is inside the boundary. The mains and the tie breakers in a mission-critical lineup are the classic place for energy-reducing maintenance switching.

Every applicable bus in the lineup gets labeled, and the labels are part of the commissioning turnover, not an afterthought added later. The critical commissioning check is that the as-left protective settings in the gear actually match the coordination and arc-flash study, because a breaker left on factory defaults blows up both the coordination and the labeled incident energy. The arc-flash and coordination work is a slice of the broader power QA scope, and it belongs in the same record set the rest of the electrical commissioning produces.

Working distance and the gear type

Incident energy is always reported at a working distance, the distance from the arc source to the worker's face and chest, and that distance changes with the gear. IEEE 1584 uses typical working distances by equipment class, and they are not the same for switchgear, an MCC, and a panelboard. The number on the label is only valid at the distance it was calculated for; stick your head closer than that and the energy you actually take is higher than the label says.

For low-voltage equipment at 600 V and below, IEEE 1584 commonly uses 18 in for panelboards and motor control centers and 24 in for switchgear. Medium-voltage switchgear typically uses 36 in. The gap between conductors and the electrode configuration shift with the gear class too, which is why the same fault current produces different energy in a panelboard than in switchgear. These are the standard's typical values; the study uses the actual gear data where it is known.

The field lesson is about posture, not just numbers. The label assumes you are standing at the working distance, square to the gear. Lean in to read a setting, reach across the bus, or kneel to land a conductor at the bottom of a tall section, and your face may be well inside the assumed distance. The PPE was rated for the labeled energy at the labeled distance. Close the gap and you have quietly underprotected yourself.

Where arc-flash programs fail in the field

The failures are not exotic. They are the same handful, and they all come from treating the study as a binder instead of a living control on the gear. The first is no study at all, a facility running on the category-method tables or on nothing, with workers guessing at gear that may be well outside the table limits.

The second is the stale study. A transformer got swapped, the service got upgraded, a breaker got replaced, and nobody re-ran the affected buses. The labels still read the old numbers, and the worst case is a label that now understates the hazard because the available fault current went up. The third is the label problem: labels missing on applicable gear, labels that mix the incident-energy and category methods so they mean nothing, or labels so faded that the boundary and the energy cannot be read.

The fourth is PPE misuse, usually the category method stretched onto gear it does not cover, or FR clothing with no arc rating treated as arc-rated. The fifth is the quiet one and the most dangerous: the breaker whose clearing time was never verified. The study assumes the device clears at a certain time, and the whole incident-energy number rests on that assumption. If the breaker was left on factory defaults, or its trip unit was never tested, the labeled energy is fiction. This is exactly the as-left settings check that electrical commissioning is supposed to catch and so often does not finish.

Field checklist

What to document

The arc-flash study is only worth what its record can prove later. The label on the gear is the field copy; the study report behind it is what an engineer needs when the system changes. Capture the inputs and the results at each bus so a reviewer can reproduce the number and a future change can be traced through it.

Record the equipment identifier, the nominal voltage, the available fault current at the bus, the protective device and its clearing time at the arcing current, the calculated incident energy and the working distance it sits at, the arc-flash boundary, the PPE result, and the date and revision of the study. Tie the as-left protective settings to the device so the next person can confirm the gear still matches the study that produced the label.

Common mistakes

• Running an arc-flash study without a current short-circuit and coordination study underneath it.
• Leaving the study stale after a transformer, breaker, setting, or utility change moved the fault current or clearing time.
• Mixing the incident-energy method and the PPE category method on the same task or label.
• Wearing PPE with an arc rating at or below the incident energy instead of above it.
• Treating flame-resistant clothing with no arc rating as arc-flash PPE.
• Stretching the category-method tables onto gear outside the parameters they cover.
• Reading a label distance while leaning well inside the assumed working distance.
• Doing energized work that could be done dead, without a justification or a permit.
• Trusting a labeled incident energy when the breaker clearing time was never verified in commissioning.
• Confusing the arc-flash boundary with the shock approach boundaries and protecting against only one hazard.

Standards and references

NFPA 70E, the Standard for Electrical Safety in the Workplace, is the worker-side framework: the arc-flash risk assessment, the boundaries, the PPE selection methods, the label contents commonly cited at 130.5(H), the choice of PPE method at 130.5(F), and the hierarchy of risk control at 110.5(H). The category tables have moved between 130.7 and 130.5 across editions, so confirm the clause against the adopted edition before citing it. IEEE 1584, the Guide for Performing Arc-Flash Hazard Calculations, is the calculation method that produces the incident energy and the arc-flash boundary from the system model.

The arc-flash study sits on top of a short-circuit study and a protective device coordination study, both built on IEEE power-system analysis practice. The NEC, NFPA 70, governs the installation, and its arc-energy reduction requirement at 240.87 mandates a means to cut clearing time on breakers rated or set at 1200 A or higher; the selective-coordination requirements live in the articles for the systems they apply to, such as emergency and legally required standby systems. The exact article numbers and their scope shift between code cycles, so confirm them against the edition the jurisdiction has adopted.

OSHA makes it enforceable. The general-industry electrical safety rules in 29 CFR 1910 Subpart S require employers to protect workers from electrical hazards, and OSHA recognizes NFPA 70E as a means of compliance. The engineer of record and the study the owner commissions govern the specific values; the standards give the framework, and the project documents and the adopted editions control.

Units, terms, and acronyms

Arc-flash work carries a vocabulary that reads differently across a study report, a label, and the standard, and the same idea shows up under more than one name. The terms below are the ones that travel across the whole subject.

Incident energy (cal/cm2)

Thermal energy a worker would receive at a set working distance from an arc, in calories per square centimeter; the study's main output

Arc-flash boundary

The distance from the gear where incident energy drops to 1.2 cal/cm2, the onset of a second-degree burn

IEEE 1584

The IEEE guide that calculates incident energy and the arc-flash boundary; the calculation method, not the safety rules

NFPA 70E

The NFPA standard for electrical safety in the workplace; the program, boundaries, PPE, permit, and label rules

Arc rating (ATPV / EBT)

The cal/cm2 protection value of arc-rated clothing, from the arc thermal performance value or energy of breakopen threshold

AR vs FR

Arc-rated has a tested cal/cm2 number; flame-resistant alone does not. Arc-flash PPE must be AR

PPE category (1 to 4)

The table-method classification with minimum arc ratings of 4, 8, 25, and 40 cal/cm2

Working distance

The distance from arc source to the worker's face and chest the incident energy is reported at; typically 18 to 36 in by gear type

Bolted vs arcing current

Bolted fault current is the dead-short value; arcing current is the lower current that actually flows through the arc

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is the difference between grounding and bonding?
• How long does data center electrical commissioning take?
• What is arc energy reduction?
• What are the arc flash PPE categories?
• What is lockout/tagout?